Code, copy, and images are every website's building blocks. Unfortunately, finding great, reasonably priced images is a challenge; that's why the "text to art" artificial intelligence tools hold so much promise. Today's post asking Is AI the Future of Web Images covers four topics:
We found the “Woman’s face and graduated background” image by andrew713 on the content sharing “Openverse” where artists and content creators from around the world share content. Content creators specify the Creative Commons license needed to use their work – in this case CC BY.
CC BY: This license allows reusers to distribute, remix, adapt, and build upon the material in any medium or format, so long as attribution is given to the creator. The license allows for commercial use.
There are six CC licenses we link in Resources at the bottom of the page. Creative Commons is an important organization we support, with a great store of CC logo swag, and a worthy cause.
We opened discussing Creative Commons because our favorite free image sites probably wouldn’t exist without the hard work pioneered by Creative Commons. Not long-ago stock images all looked the same, were controlled by a handful of powerful sites, and cost a fortune. Thankfully the Internet’s ability to break up monopolies kicked in, Creative Commons was formed, and creators from around the world found new ways to have their work recognized and supported.
We like Pexels and Picabay because their images feel spontaneous and less canned. We share a Buffer post in Resources with 21 free image websites. Buffer opens with Unsplash and Burst sites we plan to include in our image searches going forward.
AI art is also called “text to image” because typing a text query into an artificial intelligence image generator such as NightCafe produces an image. For example, we typed “woman wearing floppy hat and sunglasses” into NightCafe, Dall E, DreamStudio, and StarryAi waited about ten seconds, and their AI produced these images.
The same note about checking your AI generated image RGB profile mentioned in our Expert Tip above applies. We found fixing RGB sliders with Photoshop improved AI generated image quality, too.
The inspiration for this search comes from the American artist Alex Katz:
Trying to understand how diffusion models work will give you a headache. Here is an explanation from Ryan O'Connor on assemblyai (linked in Resources):
Diffusion Models are generative models, meaning they are used to generate data similar to the data on which they are trained. Fundamentally, Diffusion Models work by destroying training data through the successive addition of Gaussian noise and then learning to recover the data by reversing this noising process.
Since machine learning gets faster and faster as it does more and more "training" models to destroy and reconstitute images, it is reaching the "there's an app for that" stage. When Google steps in as they have with Imagen, you know AI art is about to explode. Google is cagey about when Imagen will be released, but we bet 2023 at the latest. We'd make that bet because we received our invitation to use Dall E after waiting less than twenty-four hours. Such a short invitation to use wait usually means an app is almost out of beta.
Copyright applies to work humans create, so the Copyright Office has consistently ruled not to grant copyrights to AI-generated images, as artist Stephen Thaler discovered. Thaler asked the office to reconsider their 2019 ruling that an image created by Thaler's "Creativity Machine" lacked the human authorship needed to support a copyright.
There are concerns that text to image AI could be abused as James Vincent shared in Anyone can use this AI art generator – that's the risk:
This (use of AI image generators) is essentially uncharted territory, and it’s not clear what the consequences of releasing a model like this into the wild will be. It’s easy to imagine the many malicious uses this technology could be put to, but that doesn’t mean these predictions will all come to pass.
Is AI art the future of web images? Again, we would answer yes and no. Yes, AI is the future of web images, but not because machines will replace artistic talent, inspiration, or passion. On the contrary, we’re betting artists such as Steve Thaler or our great graphic design team will use these new AI art tools to continue to inspire, motivate, and create for and with us.
What do you think? Do you think AI images are the future of online graphics? Email what you think to msmith (at) WTE.net.